Environ Monit Assess (2013) 185:2211–2220 DOI 10.1007/s10661-012-2702-1
The residues and environmental risks of multiple veterinary antibiotics in animal faeces Yan-xia Li & Xue-lian Zhang & Wei Li & Xiao-fei Lu & Bei Liu & Jing Wang
Received: 14 January 2012 / Accepted: 28 May 2012 / Published online: 13 June 2012 # Springer Science+Business Media B.V. 2012
Abstract To understand the residues and ecological risks of veterinary antibiotics (VAs) in animal faeces from concentrated animal feeding operations in northeastern China, 14 VAs were identified by high performance liquid chromatography, and the preliminary risks of six antibiotics were assessed using the hazard quotient (HQ). The investigated VAs occurred in 7.41 to 57.41 % of the 54 samples, and the levels ranged from 0.08 to 56.81 mg kg−1. Tetracyclines were predominant with a maximum level of 56.81 mg kg−1 mostly detected in pig faeces. Sulfonamides were common and detected with the highest concentration of 7.11 mg kg−1. Fluoroquinolones were more widely detected in chicken faeces rather than in pig or cow faeces, which contained the dominant antibiotic enrofloxacin. In comparison, the residue of tylosin was less frequently found. The risk evaluations of the six antibiotics revealed that tetracyclines, especially oxytetracycline, displayed the greatest ecological risk because of its high HQ value of 15.75. The results of this study imply that multiple kinds of VAs were jointly used in animal feeding processes in the study area. These Y.-x. Li (*) : X.-l. Zhang : X.-f. Lu : B. Liu : J. Wang State Key Laboratory of Water Environment Simulation, School of Environment, Beijing Normal University, 100875 Beijing, China e-mail:
[email protected] W. Li Nuclear and Radiation Safety Centre, Ministry of Environmental Protection, Beijing 100082, China
medicine residues in animal faeces may potentially bring ecological risks if the animal manure is not treated effectively. Keywords Veterinary antibiotics . Animal faeces . Risk evaluations . Ecological risk
Introduction During the last decades, numerous veterinary antibiotics (VAs) have been used globally as growth promoters and therapeutic agents in livestock production because of their positive effects (Arikan et al. 2007). In the USA, for example, the use of VAs as feeding supplements has increased by 109-fold from 1950 to 2004 (AHI 2002; Arikan et al. 2009), and 60 to 80 % were used for nontherapeutic purposes (Mellon et al. 2001). It has been reported that 5,000 tons of antibiotics was consumed as veterinary drugs and growth promoters in the European Union (Sarmah et al. 2006). Many studies have shown that antibiotics administered to livestock were usually excreted without metabolism (HallingSørensen et al. 1998; Arikan et al. 2006); for example, 70 to 90 % of tetracycline may be excreted as parent compounds via urine or faeces (Halling-Sørensen 2000; Pils and Lairo 2007). It has been reported that up to 20 mg L−1 of tetracycline was observed in animal manure (Winckler and Grafe 2000). However, unlike municipal sewage treatment plants, concentrated animal feeding operations (CAFOs) do not require additional treatments as long as the animal wastes are not discharged directly
2212
into water systems. A growing number of studies in the USA and European countries have reported the existence of VAs in animal manure, bodies of water, sediments and soils (Campagnolo et al. 2002; Hamscher et al. 2002; Hu et al. 2010; Watanabe et al. 2010). Furthermore, antibiotic resistance genes have appeared and accumulated due to long-term exposures to various antibiotics (even low levels) (Kulshrestha et al. 2004; Heuer et al. 2011), which have been tightly attributed to the animal production practices, and these genes might enter human bodies through the food chains. Moreover, the microorganism ecology may be affected by contamination from these antibiotics (Kong et al. 2006; Kleineidam et al. 2010). Certainly, the fates and risks of VAs prescribed in animal industries have been raising concern in recent years. Since the early 1990s in China, VAs have been widely used in the livestock industry (Zhao et al. 2010). Statistics from the China Chemistry Industrial Association and the China Pharmaceutical Industrial Association indicated that the weight of the raw materials used for the production of antibiotics was 210,000 tons in 2005, and of these, 97,000 tons was utilised in the animal industry. Regardless of the exact amount and how they are used, it is clear that large amounts of VAs are consumed annually in China, especially in animal feedlots. Tetracyclines (TCs), sulfonamides (SAs), fluoroquinolones (FQs) and tylosin (TYL) are the most commonly used antibiotics in animal production (Xu et al. 2007). Recently, significant levels of these antibiotics have been detected in animal manure in China (Zhao et al. 2010; Pan et al. 2011). Moreover, most antibiotics have polar functional groups and high solubilities in water. Luo et al. (2011) found that SAs frequently determined in water and sediment in the Haihe River of China are closely related to the surrounding livestock and fishpond feeding. There is a long history of applying animal manure to farmland as a low-cost, organic fertilizer in China; hence, the potential pollution risks of these antibiotics should be taken into account. As the main animal production base of China, three northeastern provinces (Heilongjiang, Jilin and Liaoning Province) supply 16.0, 13.8 and 20.1 % of the meat, eggs and cow milk to the Chinese market (Department of Rural Surveys, National Bureau of Statistics 2007). Consequently, more than 190 million tons of animal manure has been produced annually (Li et al. 2007; Department of Rural Surveys, National Bureau of Statistics 2009). Furthermore, animal production has been encouraged by local governments to ensure a stable
Environ Monit Assess (2013) 185:2211–2220
supply of animal products and farmers' income. The amount of animal manure will keep increasing in these regions, and the residues of the veterinary antibiotics will have to be taken into account. Because the types and doses of VAs administered to animals vary broadly depending on the type of animal, the growth stage, the epidemic situation and the feeding brands, residues of VAs in animal manure differ among the animal types and feedlot breeding areas examined in the literature (McEwen and Fedorka-Cray 2002; Kumar et al. 2005; Zhao et al. 2010). The present study established a method for simultaneously determining 14 commonly used antibiotics in animal manure. We then used this method to detect antibiotics in different animal faeces including pig, chicken and dairy cow faeces from CAFOs in three northeastern provinces of China. Finally, the potential risks of six VAs were evaluated to understand whether the animal faeces pose risks to the environment when they were introduced into fields by agricultural practices.
Materials and methods Field sites and sampling The area of study is 787,800 km2, located in northeastern China, and is comprised of the Heilongjiang, Jilin and Liaoning provinces (E 118° 53′–105° 05′, N 38° 43′–53° 33′). These provinces are the most important agricultural regions in China and have the highest concentrated animal production base in China. The sampling sites are distributed across nine cities in the three provinces, which represent the major animal production regions (Fig. 1). Fifty-four animal faecal samples were collected from 18 pig farms, 18 dairy cattle farms and 18 chicken farms. The sizes of the farms ranged from 100 to 10,000 animals for the pig farms, from 40 to 3,000 animals for the dairy cow farms, and from 800 to 100,000 animals for the chicken farms. At each farm, fresh, solid droppings were collected from multiple locations and mixed thoroughly. Approximately 1,000 g of mixture was collected as a representative sample at each farm. The samples were immediately put into a container, chilled at a temperature of −4 to 16 °C, delivered to a laboratory and stored in a freezer below −20 °C in the dark. All of the samples were freeze-dried, sieved through 2-mm pores and analysed within 4 weeks.
Environ Monit Assess (2013) 185:2211–2220
2213
Fig. 1 Distribution of sampling sites in the study area
Reagents and standards Target compounds were selected based on their usage in animal production and reports from literatures (Zhang et al. 2008). Tetracyclines including tetracycline (TC), oxytetracycline (OTC) and chlortetracycline (CTC); sulfonamides including sulfaguanidine (SG), sulfanilamide (SA), sulfamethoxazole (SMZ), sulfamonomethoxine (SMM), sulfamerazine (SMR) and sulfachlorpyridazine (SCP); fluoroquinolones including norfloxacin (NOR), ciprofloxacin (CIP), enrofloxacin (ENR) and difloxacin (DIF); and tylosin (TYL) were investigated in this study. Due to its uncommon use in animal production in China, lomefloxacin (LOM) was chosen as an internal standard for analysis. Standards of TC (>97.5 %), OTC (>96.0 %), NOR (>99.6) and CIP (>84.9 %) were purchased from the National Institute for the Control of Pharmaceutical and Biological Products, Beijing, China. CTC (>99.8 %),
SG (>99.9 %), SA (>99.9 %), SMZ (>99.6 %), SMM (90.0 %), SMR (>99.9 %), SCP (>99.5 %), ENR (>99.9 %), DIF (>99.8 %) and TYL (>82.4 %) were obtained from Sigma-Aldrich Chemical Company, USA. Methanol, acetonitrile and dichloromethane were of liquid chromatographic grade and obtained from Fisher Company, USA. Na2EDTA, NaOH, citric acid and acetic acid were of analytical grade. Ultrapure water was achieved using the Milli-Q Advantage A10 Water Purification System (US Millipore Co., Bedford, MA). All of the experiments were conducted in glassware that were silanised prior to use. The stock solutions of TCs, SAs and TYL were prepared by dissolving 10 mg of each antibiotic separately into 10 ml of methanol (MeOH). The stock solutions of FQs were prepared by dissolving 10 mg of antibiotic separately into 10 ml of NaOH solution (0.03 mol L−1). All stock solutions were stored at 4 °C in amber vials and were freshly prepared every 2 months. The working solution was diluted from
2214
fresh stock solution on a daily basis. Analytical quality control tests including blanks and spiked samples (spiked with individual antibiotics at 1 and 10 mg kg−1, separately) were included in each batch of samples analysed. Sample preparation and detection The animal faeces samples were extracted in an accelerated solvent extraction system (Dionex, ASE3000). Exactly 1.0 g of lyophilised faecal sample, 0.5 g of diatomite and 0.3 g of EDTA were added in sequence into an extraction tank that was covered beforehand at the bottom by a filter membrane cushion. Meanwhile, a mixture of MeOH and citrate solution (0.2 mol L−1, pH 4.00) (V/V01:1) was added into a regent vial. Samples were extracted statically over two cycles with each cycle lasting for 10 min at room temperature at 1,500 psi. The washing volume and blowing time were set at 120 % and 180 s, respectively. The extracted solution was decanted into a large vial (1 L) and diluted by ultrapure water to ensure a content of MeOH below 10 %. The dilutions were loaded on an Oasis hydrophilic– lipophilic balance cartridge (500 mg/6 ml, Waters, USA), which was activated in advance by 5 ml MeOH and 5 ml of 0.2 mol L−1 citrate solution, successively. Dichloromethane (10 ml) was applied to wash the objectives at a rate of 5 ml min−1 after being prewashed with 10 ml of methanol solution (V/V, 5:95) and 10 ml of ultrapure water. The whole process was performed in a special vacuum device (VisiprepTM-DL, Supelco Company, USA). The eluent was prepared using a nitrogen pressure blowing concentrator (WD-12, Aosheng Company, Hangzhou, China) to near dryness, adding 1 μg internal standard of LOM and adjusting to approximately 1 ml by a mixture of MeOH and ultrapure water (V/V, 2:3) and filtering it through a 0.2-μm membrane using a syringe filter for high performance liquid chromatography (HPLC) detection. The detection parameters for the 14 VAs are summarised in Table 1. The chromatograms are shown in Fig. 2. Method validation In this work, we used HPLC instead of liquid chromatography coupled with tandem mass spectrometry for detection of VAs based on the following two considerations: firstly, this study aimed to investigate the VA residues in animal faeces from CAFOs in which most residues may be high enough to be detected by
Environ Monit Assess (2013) 185:2211–2220 Table 1 Detection parameters of the 14 veterinary antibiotics Project
Parameter
HPLC
Agilent-1200, USA
Detector
PDA
Wavelength
270 nm
Column
Inertsil ODS-3 (Dikma) 250×4.6 mm, 5 μm
Column temperature
30 °C
Flow rate
0.8 ml min−1
Mobile phase
Mobile phase A, acetonitrile Mobile phase B, 1 % acetic acid, pH 2.6
Gradient elution
0 min, 10 % A; 18 min, 37 % A 20 min, 80 % A 22 min, 80 % A 25 min, 10 % A
HPLC; secondly, the expected detection limitation in this study is 100 μg kg−1 which is the critical concentration for further risk assessment, and it could be realized by HPLC. In order to further ensure the accuracy and precision of this method, detailed quality assurance measures including blanks, faeces-spiked recovery, limitation of detection and quantification, experimental repeatability and reproducibility were conducted in the development of this method. In addition, strict quality control in the analysis of field samples was also carried out in this study. An internal standard method was employed to quantify the concentrations of antibiotics in solution. Calibration curves were established with 7 points along the range of 0.1–50 mg kg−1. The correlation coefficients of the objectives were all above 0.999. Spiked samples were set at two concentrations to simulate the levels of actual samples and were analysed in four replicates. The recovery of individual antibiotics was calculated by dividing the detected concentration by the spiked concentrations in the samples. The average recovery was 79.6 % at the level of 1 mg kg−1 and 85.3 % at the level of 10 mg kg−1. Intra-repeatability and inter-reproducibility experiments were conducted according to the method recommended by Rong and Liang (2008). The relative standard deviations of both experiments were below 15 %. The limit of detection (LOD) and limit of quantification were determined for the concentration
Environ Monit Assess (2013) 185:2211–2220
2215
Fig. 2 The chromatograms of the 14 veterinary antibiotics.
*Internal standard of lomefloxacin (LOM)
of the solution at the ratio of signal to noise (S/N) to be 3 and 10, respectively (Jacobsen and Halling-Sørensen 2006). Except TCs, the LODs of all the selected antibiotics were below 100 μg kg−1.
of veterinary antibiotics that were present in half of the samples investigated. The detected concentrations of tetracyclines varied by a factor of 100. Of the three tetracyclines, chlortetracycline was used the most commonly, while oxytetracycline was administered in the largest amount for which the mean value was two and four times as large as that of tetracycline and chlortetracycline, respectively. In addition, the utilisation of tetracyclines in pig feeding farms was more popular than in chicken and cow feedlots. The sulfonamides were less commonly used than the other veterinary antibiotics in these concentrated animal feeding operations analysed in this study. Compared to tetracyclines, the presence rates and residual levels of
Results and discussion The residues of veterinary antibiotics in animal faeces One or more antibiotics were simultaneously detected in 51 of the 54 faecal samples with concentrations ranging from 0.08 to 56.81 mg kg−1 (Table 2). The tetracyclines were the most common of the four types
Table 2 Occurrence and detected concentration of 14 VAs in animal faeces VA
Range (mg kg−1)
Occurrence (%) Dairy cow (n018)
Chicken (n018)
Pig (n018)
Dairy cow (n018)
Chicken (n018)
Mean (mg kg−1) Pig (n018)
Dairy cow (n018)
Chicken (n018)
Pig (n018)
TC
33.3
44.4
50.0
0.43–2.69
0.54–4.57
0.32–30.55
1.08
1.83
5.29
OTC
27.8
38.9
50.0
0.21–10.37
0.96–13.39
0.73–56.81
5.10
6.45
11.81
CTC
55.6
50.0
61.1
0.61–1.94
0.57–3.11
0.68–22.34
1.04
1.29
3.19
SG
22.2
22.2
22.2
0.16–0.30
0.14–0.40
0.15–1.90
0.21
0.23
0.63
5.6
11.1
11.1
0.08–0.08
0.09–1.53
0.10–0.12
0.08
0.81
0.11
SMR
SA
11.1
11.1
11.1
0.10–0.11
0.14–0.89
0.13–0.15
0.11
0.52
0.14
SMZ
27.8
27.8
27.8
0.22–1.02
0.25–7.11
0.21–2.16
0.46
2.23
1.07
SMM
22.2
22.2
27.8
0.14–0.30
0.14–0.96
0.12–4.84
0.22
0.40
1.14
SCP
11.1
11.1
16.7
0.15–0.26
0.27–0.33
0.13–2.13
0.21
0.30
0.85
NOR
22.2
27.8
27.8
0.43–1.76
0.50–9.52
0.41–3.18
0.85
2.72
1.10
CIP
27.8
27.8
27.8
0.28–0.84
0.33–2.94
0.31–0.96
0.53
1.03
0.49
ENR
38.9
38.9
44.4
0.46–4.17
0.33–15.43
0.36–2.22
1.18
3.33
0.87
DIF
5.6
11.1
5.6
0.16–0.16
0.13–1.25
0.14–0.14
0.16
0.69
0.14
TYL
16.7
16.7
22.2
0.22–0.28
0.23–0.34
0.23–1.88
0.25
0.28
0.69
TC tetracycline, OTC oxytetracycline, CTC chlortetracycline, SG sulfaguanidine, SA sulfanilamide, SMZ sulfamethoxazole, SMM sulfamonomethoxine, SMR sulfamerazine, SCP sulfachlorpyridazine, NOR norfloxacin, CIP ciprofloxacin, ENR enrofloxacin, DIF difloxacin, and TYL tylosin
2216
sulfonamides were much lower. The average of the presence rates of the six investigated sulfonamides was approximately 20 %, which was half of the average percentage of the frequencies of tetracyclines. The average residual levels of sulfonamides were lower than tetracyclines by nearly one order of magnitude. The residue characteristics among the six sulfonamides varied greatly. The detectable concentrations of sulfonamides in faeces ranged from 0.08 to 7.11 mg kg−1, and most of them were less than 1 mg kg−1. Sulfamethoxazole was the predominant drug in this group with average concentrations ten times those of sulfanilamide in pig faeces and ten times those of sulfaguanidine in chicken faeces. The second leading drug was sulfamonomethoxine, primarily detected in pig faeces. Contrary to the tetracyclines, no obvious residual differences were observed for the sulfonamides among the different animal types. Similar to the sulfonamides, there was a vast discrepancy between the four fluoroquinolones displayed. Enrofloxacin, a representative drug in this group, was detected at the highest rate among the four fluoroquinolones (40 %). On the contrary, only four faecal samples were detected with difloxacin. Although observed more frequently than the sulfonamides, the detected amounts of the fluoroquinolones were equivalent with the sulfonamides and much lower than the tetracyclines. The mean concentrations of the four fluoroquinolones were in the following sequence: enrofloxacin > norfloxacin > ciprofloxacin > difloxacin. Different from the tetracyclines, the fluoroquinolones were predominantly identified in chicken faeces. The residue levels of the fluoroquinolones in pig and dairy cow faeces were comparative. The residue of tylosin in animal faeces was considerably lower compared with the other three types of veterinary antibiotics. It was found in faecal samples from four pigs, three dairy cows and three chickens. The concentrations were all below 0.4 mg kg−1 with the exception of one sample of pig faeces (1.88 mg kg−1). In an investigation in Denmark, tylosin was not detected in swine manure (Jacobsen and Halling-Sørensen 2006). The small amount of tylosin residue may be explained by two reasons: (1) the administration of tylosin in animal feeding feedlots was unpopular, and (2) tylosin was easily degraded in the environment (Jacobsen and Halling-Sørensen 2006). The residual characteristics of veterinary antibiotics in this study were consistent with the results reported by Zhang et al. (2008) in northern Zhejiang Province,
Environ Monit Assess (2013) 185:2211–2220
Hu et al. (2008) in Tianjin District, Chen et al. (2008) in Jiangsu Province, Pan et al. (2011) in Shandong Province and Zhao et al. (2010) in central and southeastern China. However, different residual amounts were observed between regions. The maximums of the tetracyclines in the present study were two times larger than those in northern Zhejiang Province (Zhang et al. 2008). The residues of the tetracyclines in cow faeces and the fluoroquinolones in chicken faeces in the present study were much lower than those in the central and southeastern provinces (Zhao et al. 2010). The maximum amounts of the four fluoroquinolones in the faeces collected from the southeastern provinces were ten times higher than the levels in the present study. Notable were the enrofloxacin concentrations in chicken faeces, which differed up to 100 times between the two studies (Zhao et al. 2010). Tetracyclines were also the primary types of veterinary antibiotics used in animal production in Korea and the USA, and most of the tetracyclines were administrated in pigs (Kim et al. 2011). However, the residual levels of tetracyclines were relatively low compared to Chinese reports. Except for a high detection of 136 and 11.9 mg kg−1 for oxytetracycline reported by Winckler et al. (2003) and Dolliver et al. (2008), respectively, the residues of tetracyclines were rarely more than 1 mg kg−1 (Kemper 2008; Karcı and Balcıoğlu 2009; Motoyama et al. 2011). According to an investigation in Australia, the amounts of ciprofloxacin and enrofloxacin in pig faeces were comparable to the present study (Martínez-Carballo et al. 2007), but the enrofloxacin amount in chicken faeces was very low in both Australia and Turkey (Martínez-Carballo et al. 2007; Karcı and Balcıoğlu 2009). These results suggested that multiple antibiotics were jointly used in animal feeding processes in these countries; furthermore, the types and doses of veterinary antibiotics varied greatly among animal species (Kemper 2008; Karcı and Balcıoğlu 2009; Motoyama et al. 2011). Effects of farm size on residual features For a better understanding of the veterinary antibiotic utilisation preferences in the study area, the animal farms were clustered into five groups to investigate the features of veterinary antibiotic residues in animal faeces. As shown in Fig. 3, on all three types of animal farms, the number of detectable antibiotics simultaneously increased with the size of concentrated animal feeding operations. For pig feeding operations, the number of
Number of antibiotics detected
Environ Monit Assess (2013) 185:2211–2220
Pig Chicken
Dairy cow
10 8 6
4 2 0
1 2 3 4 5 Scale group of CAFOs investigated
2217
with the operation size (Fig. 3). However, this increase in concentration was different from the number of antibiotics observed. The level of antibiotics observed increased slowly between the first and third groups and then changed sharply between the third and the largest group. The average concentration of veterinary antibiotics in chicken and pig faeces from the largest farm was five to ten times that of the smallest farm. This result implies that the large animal farms, with pigs >1,000, chickens >30,000 and dairy cows >300, will tend to use not only more varieties of veterinary antibiotics during operation, but also a dosage that will be significantly enhanced, thus suggesting that the residues of veterinary antibiotics in the large-scale concentrated animal feeding operations will be more problematic. When the local government encourages the rapid development of concentrated animal feeding operations, animal waste and the abuse of veterinary antibiotics resulting in ecological and environmental pollution need to be monitored. Potential risk assessment of antibiotics in animal faeces
Fig. 3 Impact of farm size on the residual veterinary antibiotics. The numbers 1 to 5 represent the scale of the CAFOs. The details are as follows: for pig feeding operations, the first group< 500, 500≤the second group≤1,000, 1,000< the third group≤2,000, 2,000
